Pre-commitment
People will restrict their own future options when they expect their future self to waver.
A commitment device is a choice made now that constrains a choice later, usually by attaching a cost to backing out. It works because people know their preferences are inconsistent over time and are willing to give up flexibility to protect a plan. Take-up is low but effects among takers can be large.
Why the plan loses at the moment of choice
How it shows up in software
Pre-commitment appears as scheduled transfers, locked savings goals, focus modes that cannot be dismissed for an hour, app blockers with a delay on unlocking, and reminders a user set for themselves. The design problem is offering the lock at the moment of resolve and making it hold at the moment of temptation.
Using it well
- Offer the commitment when motivation is high, right after a success or a decision, not when the user is already struggling.
- Make the binding real but proportionate. A cooling-off period beats an absolute lock, because an unbreakable lock stops people opting in at all.
- Show the commitment in the interface where the temptation lives, so the past self is present in the moment.
- Give a clean exit that takes time rather than money. Delay preserves autonomy; penalties invite resentment.
Where it turns manipulative
- Locking in the company's revenue while calling it the user's commitment, such as an annual plan sold as a discipline aid, inverts who the device serves.
- Devices that are hard to exit and easy to enter drift into forced continuity. If the user cannot get out in the same number of steps, it is not a commitment, it is a contract they did not read.
- Penalty-based commitments can punish people for circumstances rather than weak will, and they fall hardest on users with the least slack.
Where you have seen it
Freedom and similar blockers
A locked session cannot be ended early on the same device, so the block survives the moment it is meant to survive.
One Sec
Opening a chosen app triggers a breathing delay and asks whether the user still wants to continue, a commitment made earlier and enforced at the trigger.
Vanguard and similar auto-escalation
A saver sets contributions to rise by a fixed amount each year, committing future income before it arrives.
What the research says
- Ariely and Wertenbroch, 2002 (Psychological Science)Mixed evidence
Students allowed to set their own paper deadlines chose binding early ones and performed better than students given a single final deadline, though not as well as those given evenly spaced imposed deadlines.
Small student samples. The interesting half is that self-set deadlines beat no deadlines but lost to well designed imposed ones.
- Ashraf, Karlan and Yin, 2006 (Quarterly Journal of Economics)Well evidenced
A Philippine bank offered a savings account that locked funds until a goal or date. Take-up was around a quarter of those offered it, and balances rose substantially for those who took it.
A randomised field experiment. Note the selection: most people declined the device.
- Thaler and Benartzi, 2004 (Journal of Political Economy)Well evidenced
Save More Tomorrow let employees commit future pay rises to retirement contributions. Contribution rates rose sharply among those who joined.
Single firm origin, later adopted widely. Combines pre-commitment with a default and inertia, so the mechanisms are hard to separate.
Grades are a judgement about the evidence, not about how useful the idea is. Plenty of contested effects are still worth knowing, as long as you do not cite them as settled.